(365a) Computational Study on Fluid Behavior at the Molecular Scale: Interfacial Phenomena, Confinement Effects, and Rheology

Research Experiences

My research focuses on exploring microscopic structural and flow mechanisms of fluids at the interface, in nanoscale confinement, and under shear using computational methods ranging from molecular simulations to machine learning. The research spanned a broad spectrum of applications, including enhanced oil/gas recovery in petroleum industry, CO2 capture and storage, and the design of high-performance machine lubricants and hydraulic fracturing fluids.

1. Fluid at the interface

Fluids at the interface play a crucial role in a variety of scientific and industrial applications. For example, emulsions (water and oil mixtures) are omnipresent in cosmetics, food industry, and petroleum industry. The thickness of interfacial fluids is typically just a few molecular layers (~nm scale), making experimental observation particularly challenging under extreme conditions. Simulation methods offer an alternative yet powerful approach to explore the fluids at the interface. I have demonstrated how the heavy components in natural gas (mixtures of methane, ethane, propane, etc.) as well as the salt ions (concentration and ion type) in water affect the interfacial properties between natural gas and water. This can help optimize the operations in gas and gas hydrate production. I have advanced the understanding of surface-active components such as intrinsic polar components in oil, surfactants, and functionalized nanoparticles on oil-water interfacial properties. I have also investigated the structural and dynamical behaviors of asphaltene on the oil-water interface subjected to electric field. These studies provide new insights into the enhanced oil recovery and demulsification by electric field.

2. Confined Fluids

Fluids in nanoscale confinement exhibit unique properties and behaviors due to the significant influence of surface effects, restricted geometries, and high surface-to-volume ratios. These distinct characteristics lead to a variety of innovative applications across different fields including shale gas/oil recovery, CO2 capture and storage, and nanostructured electrodes in batteries. I have investigated CO2 distribution in the water (or brine)-filled nanopores composed by kaolinite, silica, or kerogen. The solubility of CO2 under these conditions could be enhanced or reduced compared with that under bulk conditions. Multiple factors affect CO2 distribution and solubility such as pH of water, solid surface chemistry, and pore size, which are comprehensively studied in these works. Applications of these works directly relate to the efficiency of CO2 geo-sequestration. I have also advanced the understanding of methane transport in the amorphous kerogen nanopores, emphasizing the importance of pore connectivity and tortuosity in gas diffusion.

3. Fluid Rheology

Accurate understanding of rheological properties such as shear viscosity of small-molecular and polymeric liquids is critical to optimizing their performance in applications such as lubrication and hydraulic fracturing. Shear experiments at high strain rates get affected by uncontrollable temperature rise during shear. Nonequilibrium molecular dynamics simulations have emerged as a powerful tool to extract rheological properties of fluids sheared at high strain rates that are difficult to reach in experiments. Using these simulations, I have explored the shear rheology of different small-molecular fluids spanning a wide range of molecular sizes and shapes under diverse thermodynamic conditions covering several orders of magnitude of Newtonian viscosity. I demonstrated that shear thinning of fluids under low and high Newtonian viscosities stems from distinct microscopic mechanisms differing in the contribution of molecular alignment to viscosity reduction. These mechanisms are checked by extracting the fluid structural properties and dynamical properties with the assistance of unsupervised machine learning methods. This research provides new insights into the debate of appropriate models to describe shear thinning of fluids under high pressure and strain rates and helps optimize the design of lubricants and hydraulic fracturing fluids.

Research Interests

My future research will use the aforementioned investigations as a platform to advance our in-depth understanding of fluids at the nanoscale. Experiments and molecular simulation generate a huge amount of data, but the conventional analysis methods generally under-utilize these data. I will develop and apply novel machine learning methods, including dimension reduction and deep learning based techniques to more efficiently use these data in order to deeply interrogate the link between atomic scale features and macroscopic properties. My future works will try to answer the following questions:

1. How to use machine learning to help design the surfactants under different thermodynamics conditions to control the interfacial properties? How to better simulate the emulsions at multiscale (from ~nm to ~um)? How to improve the high-fidelity simulations on the salt ions with strong polarization at the interface?
2. How to accurately predict fluid transport in nano-porous media? What key chemical and physical factors influence the fluid transport? What is the best way to represent the features of fluids and nano-porous media for use in data-driven approaches to learn the physics of these systems?
3. How to accurately predict the fluids viscosity based on the molecular features, shear rate, and thermodynamic conditions? Simulations can often take intractably long times to reach steady states or explore rare events; in such cases, how to efficiently accelerate molecular simulations via machine learning? How to use generative machine learning to quickly sample all the possible microstates of the systems? Experiments and simulations are suitable for low and high shear rates, but the intermediate shear rates are not accessible by either method; how to bridge this gap?